How a protein folds into its native structure is one of the most important and challenging problems in biological science today. A central issue lies in clarification of the early folding events, which can be very difficult to study. Several lines of evidence suggest that the initial step in protein folding involves the collapse of a polypeptide chain. However, it is unclear whether the collapse is associated with any secondary / tertiary structure formation. If structural elements form early in the folding process, are they native-like or could they contain non-native (misfolded) elements that could retard or accelerate the subsequent folding events? Are these processes different for each protein or are there general rules that are common to all proteins? For decades, folding reactions have been studied with stopped-flow instrumentation in which the typical mixing dead time is on the order of a few milliseconds during which a large portion of the reaction may be missed. Our group pioneered the development of sub-millisecond mixers for studying the early folding events. With this technique, we have been able to observe the folding of a lipid binding protein, cytochrome c and sperm whale apo-myoglobin in the submillisecond time domain for the first time. Based on these studies, we proposed a biphasic mechanism, which guarantees that the protein folds into its unique native conformation with high efficiency and fidelity. The high efficiency is made possible by a kinetically controlled nascent phase, in which the conformational space is reduced through polypeptide chain condensation; the high fidelity is achieved through the subsequent thermodynamically controlled equilibrium, in which the energy is minimized by structural fluctuations. We will further test this hypothesis in new studies on the lipid binding protein, cytochrome c and myoglobin systems. The impact of the structure of the early intermediates on the overall folding kinetics will be examined. Site directed mutagenesis will be used to create alterations in key elements involved in folding. To initiate the folding, we will exploit our well-characterized rapid mixers with dead times of 100 microseconds. In addition, new mixers fabricated from silicon have been developed for a freeze quench application with a 50 microseconds dead time. The structures of the intermediates will be studied by visible and UV resonance Raman scattering, by tryptophan fluorescence, by infrared spectroscopy and by spin labeled EPR spectroscopy. It is anticipated that this integrated approach will lead to an in depth understanding of the folding pathways of these diverse protein systems. [unreadable] [unreadable]